Phosphorus, Sulfur, and Chlorine in Fuel
Gases: Impact on High Temperature Fuel
Cell Performance and Clean-Up Options
OLGA A MARINA
Pacific Northwest National Laboratory
Workshop on Gas Clean-Up for Fuel Cell Applications
March 6-7, 2014
57% net electrical efficiency on methane 8 SOFC cells per furnace with independent gas flow
Multi-cell MCFC test stand
2
High Temperature Fuel Cell R&D at PNNL;
Impurities Overview
OA Marina
Selected Impurities in Biogas/Landfill Gas: Cell/stack/system design, fabrication,
and testing
Identification and mitigation of
performance degradation
mechanisms in cells and stacks
Development and application of
modeling and simulation tools for
improved cell/stack/system design
Partnered with industrial developers
to accelerate commercialization of
fuel cell systems
Gen 4
403 cm2 Active Area
Continuum
Electrochemistry
modeling
SOFC Component
Lamination
* SOFC and/or MCFC assessments conducted at PNNL (Marina, Pederson et al, ECS Transactions 26 (2010) 363-370)
•Class I: least volatile and will remain in the ash •Class II: more volatile and will partition between solid and gaseous species •Class III: highly volatile and show little tendency to condense from the vapor phase
Impurity VolatilityClass
*AsH3 II*HCl III*H2S II*Hg II
K INa I
*PH3 II*Sb II*Se II
Pigeaud, Maru, Wilemski, Helbe, Trace Element Emissions, DOE, 1995, pp 7-13.
Phosphorus in Fuel Gas: Very Low Threshold for
Reaction with Ni; Excellent Cleanup Options EBSD/SEM micrographs of electrolyte-
supported SOFC cells following exposure
to synthesis gas containing phosphorus
OA Marina
First nickel phosphide forms at exceptionally low
P(g) fugacities. Alkali carbonates can capture P to
below the thermodynamic threshold for Ni3P(s)
s
1 ppm, 73 hrs1 ppm, 41 hrs
1 ppm, 162 hrs1 ppm, 112 hr
30 um
Red: Ni; Green: YSZ; Blue: Ni3P
Bentonite-carbonate absorber
Phosphorus captured at inlet of fuel cell stack;
Similar behavior for arsenic and antimony in fuel gas Marina et al. SSI 181 (2010) 430; J. Power Sources, 196 (2011) 636; J. Power Sources, 195 (2010) 7140; 193 (2009) 730; J. 196 (2011) 4911.
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1.05
0.1 1 10 100 1000 10000 100000
Fra
cti
on
Alk
ali
Ca
rbo
na
te R
em
ain
ing
Time (Days)
Time (Years) 0.5 1 2 5 10 20 50
Anode Poisoning by HCl(g) Low and Reversible for Both
SOFC and MCFC; Clean-Up to <1ppm by Solid Sorbents
SOFC anode poisoning consistent with reversible surface adsorption
Ni chloride solid phases not expected
NiClx(g) activities negligibly low
Chlorides will displace carbonates in MCFCs
Electrolyte loss by chloride volatilization is the principal degradation mechanism OA Marina
0.20
0.25
0.30
0.35
0.40
0 50 100 150 200
Cu
rren
t d
en
sit
y (
A/c
m2)
Time (Hours)
HCl off
SOFC, 50 ppm HCl(g), 750oC
Chlorides as high as 1 wt% in biomass
Concentrations to 500 ppm in syngas reported
Calculated for button cell Marina et al, J. Power Sources, 195 (2010) 7033
Ni Anodes for Both SOFC and MCFC Reversibly
Poisoned by H2S(g) in Fuel Gases; Modified
Anodes Provide Higher Sulfur Tolerance
Modest decrease in fuel cell performance
due to <10 ppm H2S(g) exposure. 5
0.5
1
1.5
-50 -25 0 25 50 75 100
No
rma
lize
d A
SR
Time (Hours)
Ni
Sb/Ni
Sb
1 ppm H2S
Sn/Ni
Sn
750oC, SOFC
650oC, MCFC
750oC, SOFC
Options for improving
sulfur tolerance exist:
e.g. special alloys,
addition of ceria
High fuel utilization, high anodic polarization
can dramatically lower the concentration of S or
Se needed to form solid reaction products at
the anode/electrolyte interface OA Marina
0.5 ppm H2Se
-0.5
-0.4
-0.3
-0.2
-0.1
0
0 50 100 150 200 250 300
Time (Hours)
Vo
lta
ge
Lo
ss (
V)
0.4 A/cm2
0.2 A/cm2
0.1 A/cm2
0.25 A/cm2
700oC
V at 0 h =0.86 (0.1), 0.8 (0.2),
0.76 (0.25), 0.66 V (0.4)
Ghezel-Ayagh, ECS
Transactions, 51
(2013) 265-272.
Marina et al, JES, 158 (2011) B36-B43
Marina et al, JES, 158 (2011) B424-B429
Marina et al, JES, 158 (2011) B36-B43
Research Needs, High Temperature Fuel Cells
More thorough understanding of mechanisms of
degradation resulting from interactions with impurities in
biogas, landfill gas, shale gas, and other fuels
Long-term testing of cells and stacks under realistic
conditions of current density and fuel utilization
Expected impurity concentrations following cleanup
Expected impurity concentrations with cleanup system upset
Testing with emerging cleanup options
Thermochemical modeling of impurity interactions –
significant gaps in data exist relevant to fuel cell systems.
OA Marina
6
Phosphorus, Chlorine and Sulfur in Biomass
Phosphorus, wt % Chlorine, wt % Sulfur, wt %
Concentrations in biomass
0.14 (loblolly pine) 0.29 (ND weeds) 2-4 (grass ash) 11.5 (willow ash)
0.5 (corn straw) 0.1 (woody) 0.42-0.96 (eucalyptus on saline sites)
0.13 (corn straw) 0.03 (woody) 0.42 (dairy) 0.52 (feedlot) 1.45 (sewage)
Interactions with SOFC
Forms NixPy solid phases even below 1 ppb
<100 ppm – minor reversible losses Solid phases unlikely
<0.1 ppm – no effect 1-10 ppm –minor reversible losses >1000 ppm – solid phases formed
Interactions with MCFC
No direct study, but expect formation of NixPy solid phases even below 1 ppb
Loss of electrolyte through alkali chloride volatilization
<0.1 ppm – no effect 1-10 ppm –minor reversible losses Dissolution in molten carbonate electrolyte
March 13, 2014 7
Lack of complete agreement in literature regarding tolerance for both MCFCs and SOFCs to Cl and S